Electron paramagnetic resonance (EPR) spin-trapping has been extensively used to detect the formation of transient free radicals in biological and chemical systems. The technique uses a spin-trap that reacts with the free radical to form a persistent spin-adduct, which can be detected directly by EPR. In many cases, radical identification can be accomplished as the spin-adduct of a particular radical has unique spectral features. Spin-trapping has generally been used in a qualitative manner as the spin-adducts are usually too low in concentration and/or too unstable to allow accurate quantification. A good signal to noise ratio is essential for accurate quantification as the EPR spectrum has to be integrated twice in order to derive the concentration of free radical in the sample. We have developed a methodology whereby spectra with poor signal to noise can be quantified. This method uses a spectral simulation to generate a noise free representation of the ori gina l spectrum, which can then be quantitatively analyzed by integration. The concentration of individual spin adducts in complex mixtures can be determined. In addition the kinetics of radical production can be analyzed, as signal averaging need not be employed. This approach has been used to determine the kinetics of superoxide formation from xanthine/xanthine oxidase and nitric oxide synthase.